the seo metagame

I have put an embarrassing number of hours into Age of Empires II. Not competitively - the campaigns, mostly, and custom scenarios where I’d give myself a walled city and a bunch of starting resources. Eventually I upped the difficulty, stopped handing myself an unfair advantage, and still beat the AI factions. I was, in my own head, quite good at this game.

Then I played my first online match. I had just about finished building my first house when a handful of enemy militia walked into my town centre and started killing my villagers. I was immediately on the back foot, in a position where I couldn’t gather resources to create my own infantry to fight back, and the game was over in minutes.

I had not been beaten by a better player. I had been beaten by a build order. Something called a drush (dark age rush), which everyone else apparently knew about and I did not.

the game beyond the game

Which is roughly where Mark Williams-Cook starts his brightonSEO talk, Improve your SEO with video games and exploits. He opens with the same story, near enough: he learned a game in single player, took those lessons online, and got taken apart.

The campaign teaches you the rules. Multiplayer teaches you that the rules were never the point. Nobody online was cheating exactly. They were using what the game allowed, in ways the creators hadn’t planned for.

Gamers call this the meta, or the metagame - the game beyond the game. It’s the accumulated consensus on the best way to win: which character, which items, which build order, which opening.

You’ll often see it claimed that META stands for Most Effective Tactic Available. It’s a genuinely useful way to explain the idea, and it is entirely made up after the fact. The word is just short for metagame.

And SEO has one. You can read every word Google publishes - internal linking, heading hierarchy, keywords, all of it - and still find yourself flailing deep in the serps, whilst other sites seem to climb to the top spots overnight.

nerf, adapt, repeat

Metas do not last. When too much of the player base converges on one strategy, one of two things happens. The developers step in and weaken it - a nerf, in the jargon - or the community works out the counter, and the counter becomes the new meta until somebody counters that. Rock, paper, scissors, with patch numbers.

Either way the meta doesn’t die. It mutates. There’s no ultimate patch, no version where the game is solved and everything is balanced. There is always a best way to play. It just stops being this one.

SEO runs the same loop, and Mark’s point is that it always has. Someone finds a thing that works. It works quietly for a while, which is the profitable part. Then it gets written up, and shared, and taught. Before you know it, SEO grifters are selling it as a service, and countless people are doing it without a sliver of subtlety. Eventually Google course corrects. A core update lands, the tactic stops working, and it often takes the worst offenders down with it.

As Lily Ray puts it: it works, until it doesn’t.

there are no patch notes

When a game gets rebalanced, the developers usually tell you. They publish patch notes. Players might not like it, but you can usally see what changed, by how much, and when.

Google publishes a name and a date, then lets you know when it’s done. Sometimes a sentence about improving the quality of its results, not “self-serving listicles now land you in the naughty corner”.

So you don’t read that you’ve been nerfed. You watch your visibility plummet as you try to work out what you can pin it on other than admitting you flew too close to the sun.

And it’s retroactive. Games nerf you going forward. Google can decide that something you did in perfectly good faith two years ago, that was fine at the time, that half the industry recommended, is now the reason you’re invisible. You get penalised under rules written after the fact.

google is the dungeon master

So in reality, Google isn’t acting like a developer. It’s the Dungeon Master.

A developer ships the rules and then leaves you alone with them. A Dungeon Master sits at the table while you play. They write rules as the session goes. They roll behind a screen, so you never see the number. They can look at the clever thing you’ve spent an hour setting up and say “no, not this time”, and offer nothing further. And they aren’t running the game for you. They’re running it for everyone at the table, and for the story, and those are not always the same thing as your character winning.

There’s a second thing a Dungeon Master has that a rulebook doesn’t, though: an agenda of their own. In commercial games, that agenda is money. Mark uses Call of Duty to make the point. A studio has two jobs pulling in opposite directions: make the game good enough that people keep playing, and monetise it hard enough that it keeps earning. Lean too far into the second and players walk. Ignore it entirely and there’s no game next year. The balance point isn’t where players would put it, and it isn’t where the finance team would put it either. It sits somewhere in between, and it moves.

Google is running the same balancing act. It needs a results page good enough that you keep coming back, and profitable enough to fund the company behind it - and the company behind it sells ads. Those two aims aren’t opposed, exactly. A useless SERP sells nothing. But they aren’t the same aim either, and every time the ad block grows, or an answer keeps you on the page instead of sending you onwards, that’s the balance being adjusted. Not against you. Just not for you.

That’s the relationship. Not adversarial - a good DM isn’t out to get you, and neither is Google - but opaque, self-interested, and permanently open to revision.

Almost every SEO frustration you’ve had is downstream of that. Not that the rules are hard. That they’re unwritten, unevenly applied, written by somebody with their own stake in the outcome, and liable to change between when you started the work and when you finished it.

the only winning move

So how do you play against a Dungeon Master?

Not by hunting for the patch notes, because there aren’t any. You play by probing. You try things, you watch what happens, and you keep records good enough to tell the difference between a ruling that changed and a coincidence that didn’t.

In practice that’s unglamorous. Test on a subset before you roll out across a whole site. Make changes small enough that you can tell which one moved the needle, and reversible enough that you can put them back. Keep dated notes of what you changed, when, and what you expected, so that in six months you have something better than a hunch. Watch your own logs rather than somebody else’s case study, because the only site whose rules you can actually observe is yours.

None of that is a trick. It’s just the difference between playing the game and reading about it.

There’s a second trap worth naming, too, because it isn’t the exploit one. Most of what gets called “the meta” in SEO isn’t an exploit at all. It’s consensus: the same title format, the same 1,500-word structure, the same FAQ block at the bottom, because that’s what the tool suggested and what everyone else did. That never gets nerfed. It just quietly stops distinguishing you from the other nine results, which is worse in its way, because there’s no graph falling over to tell you it happened. You simply never rise.

The only winning move, it turns out, is to keep playing. And to take notes.

a footnote about foam swords

I went looking for the origin of “nerfed” while writing this. It comes from a game whose designer reduced the effectiveness of swords, prompting players to complain that they were now swinging nerf swords, after the brand of foam toy. The insult stuck, got adopted by the developers themselves, and is now the standard industry term for making something worse on purpose.

Mark Williams-Cook’s talk, Improve your SEO with video games and exploits, was given at brightonSEO in April 2025. Watch it on YouTube.